water demand and availability has to be examined on a smaller scale by river basins and even subbasins. Obviously, water is in short supply in the dry and arid parts of the world, some of which are rather heavily populated. Floods constitute the main problem in other places where agricultural land and cities are subjected to devastation from excessive runoff. The problem boils down to one of adjustment of population and economic activities to the hydrologic situation in various places. A number of possibilities exist for augmenting water supplies. Additional storage reservoirs can be constructed; evaporation and irrigation canal losses can be checked; water-consuming trees and plants can be removed; water can be recycled in industry; salt or brackish water can be substituted for fresh water in cooling and certain other uses; water prices can be raised to curtail consumption; whole river systems can be interconnected; surface and ground water sources can be integrated for more economic use; and so on. In many streams pollution abatement would yield large amounts of higher quality and, therefore, more usable water than is now available. Large gains can be made by legal and institutional changes which would result in some reallocation of water use away from irrigation, which in the United States comprises more than 90 percent of total withdrawals from streams, toward industrial and other much higher value uses. Possibilities for desalinating water exist, but costs are still so high at present that economic use cannot be expected in more than a very few places. It appears that the major water problem in most places for the future will be maintaining and improving its quality, especially in densely populated metropolitan and industrialized regions. THE PROBLEM OF QUALITY We have seen that the known quantities of natural resources in the world are sufficient to feed the projected populations to the year 2000, if the requisite changes, particularly in agricultural methods, can be brought about. While some of our assumptions may be on the liberal side, others are conservative; for example, we have made no allowance for the increase in land that may become available for food production when machinery is substituted for animals as a source of power. Nor have we mentioned increased yields of food per food-producing animal—a source of large gains in the meat, egg, and milk-producing industries of the United States in the last few decades. However, the changes we have assumed require the use of large quantities of fertilizers, pesticides, and—in many areas—water to bring output per hectare up to United States and European levels of output. These changes will produce burdens on the natural environment which are potentially quite serious. The "quantity" problem may be solved at the cost of raising difficulties with "quality."s 143-148), although in most cases they appeared to be incomplete.